86 Mendenhall — Free Penduhim as a Time Standard. 



established astronomical observatory, or directly by the use of 

 a transit instrument. In the former, rates are carried during 

 bad weather and often for many days, by one or more clocks 

 and with greater or less accuracy but the signals as received 

 are subject to uncertainty and error arising out of telegraph 

 transmission. In the latter, a few nights in succession of 

 cloudy weather prevent the observer from getting a rate for 

 his time-piece just when he most needs it. In both cases the 

 daily rate is usually all that is known, although this may differ 

 widely from that existing when the particular experiment was 

 made for which the time is to be standardized. It is believed 

 that a free pendulum, vibrating under constant conditions 

 furnishes a much more reliable standard for short intervals 

 than any clock or chronometer and that this standard, may be 

 easily utilized by methods about to be described. 



The pendulum* is for convenience, a half second's pendulum 

 and is, therefore, about a quarter of a meter long. Its mass is 

 only a trifle greater than a kilogram and the most of this 

 is concentrated in the bob. The knife edge, rigidly attached 

 to the pendulum, is of agate and it swings upon agate planes. 

 These should be rigidly supported and may well be secured to 

 a part of the casting which furnishes at once the support for 

 the pendulum and the chamber in which it swings. It should be 

 furnished with a starting and stopping apparatus and an arc for 

 measuring the amplitude of its vibrations. If it is to be used 

 only for comparing chronometers and clocks no arrangements 

 for securing constant pressure and uniform temperature are 

 necessary. If it is to be a time standard both should be pro- 

 vided. The first is easy and the second can generally be 

 reached quite closely. Some method of knowing the temper- 

 ature of the pendulum should be provided. When used 

 under conditions of nearly constant temperature a " dummy" 

 pendulum, in the same enclosure, with a thermometer properly 

 embedded in its stem, will be sufficient. Its amplitude should 

 not be greater than two degrees at first, and when the knife 

 edge is properly prepared it will swing for a couple of hours 

 without this being inconveniently reduced. 



The period of the pendulum may be ascertained by means 

 of a modified coincidence method, the elements of which were 

 first applied by Herr Yon Sternek in his use of short pendu- 

 lums in gravity determinations. 



A small mirror is placed in a vertical plane on the pendu- 

 lum head, its center being as near as may be in a horizontal 



* The pendulums used in the recent gravitation work of the U. S. Coast and 

 Geodetic Survey will be found figured and described in an Appendix to the Report 

 for 1S91. 



